X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/63f861861578b33416a0b2d801252b27443fabde..25cb0dced028c91d9a73d7faedae4d3e73162902:/src/search/views.py diff --git a/src/search/views.py b/src/search/views.py index 70a216e3b..39284c70c 100644 --- a/src/search/views.py +++ b/src/search/views.py @@ -3,20 +3,22 @@ # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. # from django.conf import settings +from django.http.response import HttpResponseRedirect from django.shortcuts import render_to_response from django.template import RequestContext from django.views.decorators import cache from django.http import HttpResponse, JsonResponse -from django.utils.translation import ugettext as _ -from catalogue.utils import split_tags from catalogue.models import Book, Tag -from pdcounter.models import Author as PDCounterAuthor, BookStub as PDCounterBook -from search.index import Search, SearchResult +from pdcounter.models import Author +from picture.models import Picture +from search.index import Search, SearchResult, PictureResult from suggest.forms import PublishingSuggestForm import re import json +from wolnelektury.utils import re_escape + def match_word_re(word): if 'sqlite' in settings.DATABASES['default']['ENGINE']: @@ -25,11 +27,11 @@ def match_word_re(word): return "[[:<:]]%s[[:>:]]" % word -query_syntax_chars = re.compile(r"[\\/*:(){}]") +query_syntax_chars = re.compile(r"[\\/*:(){}?.[\]+]") def remove_query_syntax_chars(query, replace=' '): - return query_syntax_chars.sub(' ', query) + return query_syntax_chars.sub(replace, query) def did_you_mean(query, tokens): @@ -64,30 +66,31 @@ def hint(request): if len(prefix) < 2: return JsonResponse([], safe=False) - prefix = remove_query_syntax_chars(prefix) + prefix = re_escape(' '.join(remove_query_syntax_chars(prefix).split())) try: limit = int(request.GET.get('max', '')) except ValueError: - limit = -1 + limit = 20 else: if limit < 1: - limit = -1 + limit = 20 + authors = Tag.objects.filter( + category='author', name_pl__iregex='\m' + prefix).only('name', 'id', 'slug', 'category') data = [ { 'label': author.name, - 'category': _('author'), 'id': author.id, 'url': author.get_absolute_url(), } - for author in Tag.objects.filter(category='author', name__iregex='\m' + prefix)[:10] + for author in authors[:limit] ] if len(data) < limit: data += [ { - 'label': '%s, %s' % (b.title, b.author_unicode()), - 'category': _('book'), + 'label': b.title, + 'author': b.author_unicode(), 'id': b.id, 'url': b.get_absolute_url() } @@ -104,11 +107,6 @@ def hint(request): @cache.never_cache def main(request): query = request.GET.get('q', '') - query = ' '.join(query.split()) - # filter out private use characters - import unicodedata - query = ''.join(ch for ch in query if unicodedata.category(ch) != 'Co') - if len(query) < 2: return render_to_response( 'catalogue/search_too_short.html', {'prefix': query}, @@ -117,30 +115,50 @@ def main(request): return render_to_response( 'catalogue/search_too_long.html', {'prefix': query}, context_instance=RequestContext(request)) - query = remove_query_syntax_chars(query) + query = prepare_query(query) + pd_authors = search_pd_authors(query) + books = search_books(query) + pictures = search_pictures(query) + suggestion = u'' - words = query.split() - if len(words) > 10: - query = ' '.join(words[:10]) + if not (books or pictures or pd_authors): + form = PublishingSuggestForm(initial={"books": query + ", "}) + return render_to_response( + 'catalogue/search_no_hits.html', + { + 'form': form, + 'did_you_mean': suggestion + }, + context_instance=RequestContext(request)) - search = Search() + if not (books or pictures) and len(pd_authors) == 1: + return HttpResponseRedirect(pd_authors[0].get_absolute_url()) - tags = search.hint_tags(query, pdcounter=True, prefix=False) - tags = split_tags(tags) + return render_to_response( + 'catalogue/search_multiple_hits.html', + { + 'pd_authors': pd_authors, + 'books': books, + 'pictures': pictures, + 'did_you_mean': suggestion + }, + context_instance=RequestContext(request)) - results_parts = [] +def search_books(query): + search = Search() + results_parts = [] search_fields = [] + words = query.split() fieldsets = ( - (['authors'], True), - (['title'], True), - (['metadata'], True), - (['text', 'themes_pl'], False), + (['authors', 'authors_nonstem'], True), + (['title', 'title_nonstem'], True), + (['metadata', 'metadata_nonstem'], True), + (['text', 'text_nonstem', 'themes_pl', 'themes_pl_nonstem'], False), ) - for fieldset, is_book in fieldsets: - search_fields += fieldset - results_parts.append(search.search_words(words, search_fields, book=is_book)) - + for fields, is_book in fieldsets: + search_fields += fields + results_parts.append(search.search_words(words, search_fields, required=fields, book=is_book)) results = [] ids_results = {} for results_part in results_parts: @@ -151,12 +169,12 @@ def main(request): else: results.append(result) ids_results[book_id] = result - + descendant_ids = set( + Book.objects.filter(id__in=ids_results, ancestor__in=ids_results).values_list('id', flat=True)) + results = [result for result in results if result.book_id not in descendant_ids] for result in results: search.get_snippets(result, query, num=3) - suggestion = u'' - def ensure_exists(r): try: return r.book @@ -164,25 +182,61 @@ def main(request): return False results = filter(ensure_exists, results) + return results - if not results: - form = PublishingSuggestForm(initial={"books": query + ", "}) - return render_to_response( - 'catalogue/search_no_hits.html', - { - 'tags': tags, - 'prefix': query, - 'form': form, - 'did_you_mean': suggestion - }, - context_instance=RequestContext(request)) - return render_to_response( - 'catalogue/search_multiple_hits.html', - { - 'tags': tags['author'] + tags['kind'] + tags['genre'] + tags['epoch'] + tags['theme'], - 'prefix': query, - 'results': results, - 'did_you_mean': suggestion - }, - context_instance=RequestContext(request)) +def search_pictures(query): + search = Search() + results_parts = [] + search_fields = [] + words = query.split() + fieldsets = ( + (['authors', 'authors_nonstem'], True), + (['title', 'title_nonstem'], True), + (['metadata', 'metadata_nonstem'], True), + (['themes_pl', 'themes_pl_nonstem'], False), + ) + for fields, is_book in fieldsets: + search_fields += fields + results_parts.append(search.search_words(words, search_fields, required=fields, book=is_book, picture=True)) + results = [] + ids_results = {} + for results_part in results_parts: + for result in sorted(PictureResult.aggregate(results_part), reverse=True): + picture_id = result.picture_id + if picture_id in ids_results: + ids_results[picture_id].merge(result) + else: + results.append(result) + ids_results[picture_id] = result + + def ensure_exists(r): + try: + return r.picture + except Picture.DoesNotExist: + return False + + results = filter(ensure_exists, results) + return results + + +def search_pd_authors(query): + pd_authors = Author.objects.filter(name__icontains=query) + existing_slugs = Tag.objects.filter( + category='author', slug__in=list(pd_authors.values_list('slug', flat=True))) \ + .values_list('slug', flat=True) + pd_authors = pd_authors.exclude(slug__in=existing_slugs) + return pd_authors + + +def prepare_query(query): + query = ' '.join(query.split()) + # filter out private use characters + import unicodedata + query = ''.join(ch for ch in query if unicodedata.category(ch) != 'Co') + query = remove_query_syntax_chars(query) + + words = query.split() + if len(words) > 10: + query = ' '.join(words[:10]) + return query